Three Years In: Still Here, Still Building
I've been using Make.com (formerly Integromat) since 2021. I've built seven live tools on the platform, run hundreds of scenarios, spent thousands of dollars on operations. I've loved it. I've hated it. But I keep coming back.
This is my honest review. Not a sales pitch. I'll tell you what genuinely works, what genuinely frustrates, and where it stands relative to competitors.
What Make.com Does Well
The visual canvas is excellent. Make.com's interface for building scenarios is genuinely good. Each module is a box. You connect them with lines. The flow is obvious. Even complex scenarios stay readable. Compared to other platforms (I'll get to that), this is Make.com's strongest feature.
Conditional logic feels natural. Routers, filters, and branches are straightforward. If you want to route differently based on a condition, you drag in a router and set up the conditions. It feels intuitive. This is critical for complex automations.
The operation model makes sense. You pay for operations (a measure of executions). One execution of a scenario = 1-10 operations depending on modules used. This is transparent and scales naturally. More usage = more cost, but the growth is predictable.
Integration breadth is real. Make.com has integrations with hundreds of services. If you need to connect two services, it's probably supported. And if it's not, the HTTP module gets you there.
Scenarios are reliable. I've had scenarios running continuously for over a year without issues. Error handling works. Retries work. Webhooks are stable. I trust Make.com not to randomly drop data.
Where It Frustrates
The learning curve is steeper than it should be. I've been a software engineer for a decade. I still found Make.com confusing at first. Understanding modules, bundles, arrays, how data flows between modules—it takes time. For absolute beginners, the curve is steep.
Debugging is weak. When something breaks, Make.com shows you the error message from the failed module. But context is often missing. You don't get a detailed execution log or step-by-step data flow. I've spent hours hunting down a single error that would be obvious with better logging.
UI bugs exist. The interface occasionally glitches. Modules don't load, connections drop, the scenario view freezes on large scenarios. These are rare, but when they happen, they're infuriating. It's a tool used by thousands, so bugs should be even rarer.
Performance optimization is manual. If a scenario is slow, you have to figure out why. There's no profiler, no "slow steps" view. You're left guessing. Is it the HTTP request? The data transform? The conditional logic? You have to remove pieces and see.
Data transformation is limited. The native text/data manipulation tools are weak. For serious transformations, you end up calling OpenAI or writing a custom HTTP integration. It feels like Make.com skipped this crucial layer.
The pricing page is confusing. You're charged per operation, but the operation count for each module isn't documented. You have to test or guess. I've been surprised by bills because a module I thought was cheap actually costs a lot.
Real Pricing Breakdown
Let me be concrete about costs. Make.com's free tier is generous: 1,000 operations per month. That's enough for light automation.
Paid tiers: $9.99/month (10,000 ops), $18.99/month (50,000 ops), $29.99/month (100,000 ops), $99/month (200,000 ops), and higher.
For my seven tools combined, I run roughly 150,000 operations per month. That puts me at the $29.99/month tier with overage charges. My actual Make.com bill is around $60-80/month when overages are included.
That's reasonable. But here's the gotcha: I initially underestimated how many operations each tool would need. I thought Shadow Hound (my most popular tool) would cost $10/month. It actually costs $30/month because I underestimated operation counts.
The lesson: test your volume estimates. Start on the free tier, build your scenarios, and run them for a week to get real numbers.
The Learning Curve: Real Talk
If you're new to automation, Make.com will take 2-4 weeks to feel comfortable. You need to understand:
- Modules and how they work
- Bundles (how data flows through the scenario)
- Mappings (connecting one module's output to another's input)
- Conditional logic (routers, filters)
- Arrays and aggregation
- Error handling
The good news: you don't need all of these to build useful scenarios. You can start with the basics (webhooks, a couple of modules, basic routing) and be productive immediately.
I spent my first month building small scenarios, debugging them, rebuilding. By month two, I was building production-grade tools. By month three, I was optimizing for cost and speed.
If you're not willing to spend a few weeks learning, Make.com might not be for you. But if you invest the time, you can build almost anything.
Versus Zapier and n8n
Make.com vs. Zapier: Zapier is simpler and more beginner-friendly. But it's limited. You can't do complex conditional logic easily. You can't build sophisticated data transformations. And it's expensive for heavy users (I'd spend $200+/month on Zapier for what I do). Make.com is more powerful and cheaper at scale. I'd choose Make.com every time.
Make.com vs. n8n: n8n is open-source, self-hosted, and more technical. If I were building for a team and had DevOps resources, I might choose n8n. But as a solo builder, Make.com is simpler. No infrastructure to maintain. Zapier-like simplicity with Make.com-like power. That's the sweet spot.
If I were starting over, I'd pick Make.com again. It's the best balance of power, cost, and ease for what I do.
The Honest Verdict
Make.com is genuinely good. It's not perfect. The UI occasionally glitches. The learning curve is steep. Debugging is frustrating. But it's the most capable no-code platform I've used, and it's cheaper than Zapier at scale.
If you're building tools for yourself or others, if you need to automate workflows that are too complex for IFTTT or Zapier, if you value control and cost efficiency, Make.com is the move.
If you're an absolute beginner looking to automate a single workflow with a popular app, Zapier is simpler. If you're building for an enterprise with DevOps support, n8n might be better. But for most solo builders and small teams? Make.com is the best platform available.
Would I recommend it? Yes. Would I use it for everything? No. Is it worth learning? Absolutely.